Skip to main content
U.S. flag

An official website of the United States government

Official websites use .gov
A .gov website belongs to an official government organization in the United States.

Secure .gov websites use HTTPS
A lock ( ) or https:// means you’ve safely connected to the .gov website. Share sensitive information only on official, secure websites.

The Impact of Data Quality on Maintenance Work Order Analysis: A Case Study in Historical HVAC Maintenance Work Orders

Published

Author(s)

Anna Conte, Coline Bolland, Lynn Phan, Michael Brundage, Thurston Sexton

Abstract

Historical data from maintenance work orders (MWOs) is a powerful source of information to improve maintenance decisions and procedures. However, data quality often impacts an analyst's ability to calculate important Key Performance Indicators (KPIs) and analyze trends within a facility. Data quality can refer to the amount of missing data, accuracy of data, or availability of appropriate data fields. When data quality is low, analysis accuracy is reduced --- often in hidden ways. We recommend several strategies to investigate data quality issues. First, the end goal of an analysis (e.g. calculated KPIs) should dictate data quality requirements, and therefore, quality reduction investigation. Second, basic Exploratory Data Analysis (EDA) techniques can be powerful tools to discover signs of quality reduction. We contextualize these techniques to the maintenance management domain with examples. By applying these strategies, this paper illustrates the impact of data quality on KPI calculation through a case study involving a representative MWO dataset. Since analysts rarely have access to ''baseline" high-quality data (i.e., to compare against), we develop a technique based on Survival Analysis that corrects for an observed quality issue (MWO completion-date entries). We use this synthesized baseline to investigate the impacts of missing data on KPI calculations. Specifically, we motivate the need for further investigation into the use of natural language text fields to find common human errors, which we estimate to be highly non-random and cause large errors in KPI calculation, even with large sample sizes. This further demonstrates the need for of Technical Language Processing (TLP) tools to discover data quality issues, along with more community-driven techniques to perform quality corrections when discovered.
Proceedings Title
6th European Conference of the Prognostics and Health Management Society 2021
Conference Dates
June 28-July 2, 2021
Conference Location
Turin, IT

Keywords

maintenance, natural language processing, technical language processing, key performance indicators

Citation

Conte, A. , Bolland, C. , Phan, L. , Brundage, M. and Sexton, T. (2021), The Impact of Data Quality on Maintenance Work Order Analysis: A Case Study in Historical HVAC Maintenance Work Orders, 6th European Conference of the Prognostics and Health Management Society 2021 , Turin, IT, [online], https://tsapps.nist.gov/publication/get_pdf.cfm?pub_id=932288 (Accessed February 29, 2024)
Created July 2, 2021, Updated December 20, 2022